AI Creates Fake Obama

[Kontinuum]

Artificial intelligence software could generate highly realistic fake videos of former president Barack Obama using existing audio and video clips of him, a new study [PDF] finds.

Such work could one day help generate digital models of a person for virtual reality or augmented reality applications, researchers say.

Computer scientists at the University of Washington previously revealed they could generate digital doppelgängers of anyone by analyzing images of them collected from the Internet, from celebrities such as Tom Hanks and Arnold Schwarzenegger to public figures such as George W. Bush and Barack Obama. Such work suggested it could one day be relatively easy to create such models of anybody, when there are untold numbers of digital photos of everyone on the Internet.

The researchers chose Obama for their latest work because there were hours of high-definition video of him available online in the public domain. The research team had a neural net analyze millions of frames of video to determine how elements of Obama's face moved as he talked, such as his lips and teeth and wrinkles around his mouth and chin.

In an artificial neural network, components known as artificial neurons are fed data, and work together to solve a problem such as identifying faces or recognizing speech. The neural net can then alter the pattern of connections among those neurons to change the way they interact, and the network tries solving the problem again. Over time, the neural net learns which patterns are best at computing solutions, an AI strategy that mimics the human brain.

In the new study, the neural net learned what mouth shapes were linked to various sounds. The researchers took audio clips and dubbed them over the original sound files of a video. They next took mouth shapes that matched the new audio clips and grafted and blended them onto the video. Essentially, the researchers synthesized videos where Obama lip-synched words he said up to decades beforehand.

The researchers note that similar previous research involved filming people saying sentences over and over again to map what mouth shapes were linked to various sounds, which is expensive, tedious and time-consuming. In contrast, this new work can learn from millions of hours of video that already exist on the Internet or elsewhere.

One potential application for this new technology is improving videoconferencing, says study co-author Ira Kemelmacher-Shlizerman at the University of Washington. Although teleconferencing video feeds may stutter, freeze or suffer from low-resolution, the audio feeds often work, so in the future, videoconferencing may simply transmit audio from people and use this software to reconstruct what they might have looked like while they talked. This work could also help people talk with digital copies of a person in virtual reality or augmented reality applications, Kemelmacher-Shlizerman says.

The researchers note their videos are currently not always perfect. For example, when Obama tilted his face away from the camera in a target video, imperfect 3-D modeling of his face could cause parts of his mouth to get superimposed outside the face and onto the background.

In addition, the research team notes their work did not model emotions, and so Obama's facial expressions in the output videos could appear too serious for casual speeches or too happy for serious speeches. However, they suggest that it would be interesting to see if their neural network could learn to predict emotional states from audio to produce corresponding visuals.

The researchers were careful to not generate videos where they put words in Obama's mouth that he did not at some other time utter himself. However, such fake videos are “likely possible soon,” says study lead author Supasorn Suwajanakorn, a computer scientist at the University of Washington.

However, this new research also suggests ways to detect fake videos in the future. For instance, the video manipulation the researchers practiced can blur mouths and teeth. “This may be not noticeable by human eyes, but a program that compares the blurriness of the mouth region to the rest of the video can easily be developed and will work quite reliably,” Suwajanakorn says.

The researchers speculated that the link between mouth shapes and utterances may be to some extent universal for people. This suggests that a neural network trained on Obama and other public figures could be adapted to work for many different people.

http://spectrum.ieee.org/tech-talk/r...tes-fake-obama

[Acidbaron]

It's not there yet but once it is fully at its point is going to be even worse to see what's fake and what is real. Although common sense will in most cases prevail there are plenty of people who would buy into it, however those people generally also buy into said things when even just read about somewhere since it fits their world view.

I'm not sure what can come out of this technology that is actually good for us, perhaps more AI based teacher programs to aid students further where there's more interaction between the user and the program? Education and entertainment seems to me to only good places it will be used.

Also i don't think the technology will be at that point so soon that we can't spot faults in the AI, however soon might have meant in this article a few years from now. I do see this doing more harm than good especially at the start, let's say someone said something actually bad or incriminating, can one than not claim it is faked? Other way around works also people being put words in their mouth.

[Wyrt]

This technology will destroy civilization.

[zenkai]

A racist AI that blames Bush for all it's shortcomings? That's some scary shit.

[Deleted]

At the rate technology is advancing, I have no doubt Trump will soon release an army of Trump-bots.

[Cerilis]

At the rate technology is advancing, I have no doubt Trump will soon release an army of Trump-bots.

They are already here!

[Ghostpanther]

Easy to do with someone like Obama. He was a master of deception.

[Deleted]

They are already here!

No, I mean like actual doom-bot, physical robots only with majestic blonde manes

[helebelemon]

It's not there yet but once it is fully at its point is going to be even worse to see what's fake and what is real. Although common sense will in most cases prevail there are plenty of people who would buy into it, however those people generally also buy into said things when even just read about somewhere since it fits their world view.

I'm not sure what can come out of this technology that is actually good for us, perhaps more AI based teacher programs to aid students further where there's more interaction between the user and the program? Education and entertainment seems to me to only good places it will be used.

Also i don't think the technology will be at that point so soon that we can't spot faults in the AI, however soon might have meant in this article a few years from now. I do see this doing more harm than good especially at the start, let's say someone said something actually bad or incriminating, can one than not claim it is faked? Other way around works also people being put words in their mouth.

Read this article as shared from my Adobe colleagues saying my alma mater is slow and given that they're preparing VoCo for release, what Adobe has implemented is very likely well beyond what the UW has created, and those concerns are definitely valid ones likely to be realized. I especially with the bolded as there is going to be a lot of bad coming out of such technology. Unfortunately, such a use case makes such technology quite futile as most entertainment shops probably can and preferably probably want to hire the real thing over attempting to spoof them, while indie shops and people would benefit the most.

[Solarnova]

24 Season 2 dealt with 'voice morphing' in which the entire premise of the second half of the season was based on a lie, due to voice manipulation on a convincing scale. It's one thing when it's a falsified voice recording, it's a whole other thing with convincing video, this will have a self-destructive effect on our civilization, when world leaders are 'morphed' and say things that they never actually said and the masses believe the false recording.

[Endus]

24 Season 2 dealt with 'voice morphing' in which the entire premise of the second half of the season was based on a lie, due to voice manipulation on a convincing scale. It's one thing when it's a falsified voice recording, it's a whole other thing with convincing video, this will have a self-destructive effect on our civilization, when world leaders are 'morphed' and say things that they never actually said and the masses believe the false recording.

That'll last up until people stop being so gullible.

It used to be that people thought "if it's printed in the news, it MUST be real!" We're pretty clearly past that, right? Same deal.

[Manabomb]

Easy to do with someone like Obama. He was a master of deception.

Ahahahahah ahaha. Hahaa.

I guess that makes Trump "The Master of Disguise"?

[Deleted]

No, I mean like actual doom-bot, physical robots only with majestic blonde manes

Getting rid of CNN, one body slam at a time.

[Deleted]

While i am all for technical advance, this could get us towards 1984.

[PC2]

It's not there yet but once it is fully at its point is going to be even worse to see what's fake and what is real. Although common sense will in most cases prevail there are plenty of people who would buy into it, however those people generally also buy into said things when even just read about somewhere since it fits their world view.

I'm not sure what can come out of this technology that is actually good for us, perhaps more AI based teacher programs to aid students further where there's more interaction between the user and the program? Education and entertainment seems to me to only good places it will be used.

Also i don't think the technology will be at that point so soon that we can't spot faults in the AI, however soon might have meant in this article a few years from now. I do see this doing more harm than good especially at the start, let's say someone said something actually bad or incriminating, can one than not claim it is faked? Other way around works also people being put words in their mouth.

It may be that the practical applications for this are smaller than the problems created. But the research for generative models will be done either way because the brain is also generative and that is the larger goal. A lot of our intelligence comes from the ability to take some real experiences and then experiment with fiction.

[Evil Midnight Bomber]

24 Season 2 dealt with 'voice morphing' in which the entire premise of the second half of the season was based on a lie, due to voice manipulation on a convincing scale. It's one thing when it's a falsified voice recording, it's a whole other thing with convincing video, this will have a self-destructive effect on our civilization, when world leaders are 'morphed' and say things that they never actually said and the masses believe the false recording.

Was also used in Tango and Cash.

[ro9ue]

Seems like we'll just have to a) rely on our sources and b) maybe have some minor 'verified' badge (like Twitter).

So if I find some crazy outlandish speech from Obama like this from a strange internet source, but The New York Times, CNN or other main media outlets aren't verifying it, I wouldn't trust it.

[ghotihook]

That'll last up until people stop being so gullible.

It used to be that people thought "if it's printed in the news, it MUST be real!" We're pretty clearly past that, right? Same deal.

So basically you're saying "we're fucked"

[Lollis]

I imagine that this is just a more advanced version of how lip sync is done in some source games; it uses syllables and produces a lip shape with them. This'll be detecting the letters, just like a speech to text program works, and then filtering it through their shape engine which renders it onto a face.

[Blood Raven]

No, I mean like actual doom-bot, physical robots only with majestic blonde manes

Trump-formers! Robots in disguise!